Structural and Functional Characterization of an Ancient Bacterial

Aug 31, 2015 - Additional experimental procedures, detailed description of the molecular dynamics simulations methods and results, ClustalW alignment ...
0 downloads 3 Views 2MB Size
Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Structural and functional characterization of an ancient bacterial transglutaminase sheds light into the minimal requirements for protein cross-linking Catarina G Fernandes, Diana Plácido, Diana Lousa, José A. A Brito, Anabela Isidro, Claudio Manuel Soares, Jan Pohl, Maria Arménia Carrondo, Margarida Archer, and Adriano O. Henriques Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.5b00661 • Publication Date (Web): 31 Aug 2015 Downloaded from http://pubs.acs.org on September 3, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Structural and functional characterization of an ancient bacterial transglutaminase sheds light into the minimal requirements for protein cross-linking Catarina G. Fernandes†,±,◊, Diana Plácido‡,±,◊, Diana Lousa§,, José A. Brito‡,, Anabela Isidro† ,◊, Cláudio M. Soares§, Jan Pohl¤, Maria A. Carrondo‡, Margarida Archer‡ ,*, and Adriano O. Henriques† ,* This work was funded by “Fundação para a Ciência e a Tecnologia” (FCT) trough PhD (SFRH/BD/43200/2008, SFRH/BD/14384/2003, SFRH/BD/28269/2006) and post-doctoral (SFRH/BPD/8967/2002, SFRH/BPD/79224/2011) fellowships and FCT Grants (PestE/EQB/LA0004/2011, POCI/BIA-BCM/60855/2004, POCTI/BCI/48647/2002, PTDC/BIAPRO/118535/2010). Support from the European Commission under the 6th Framework Programme through the Key Action “Strengthening the European Research Area Programme” (Contract nº RII3-CT-2004-506008) was provided to collect X-ray data at EMBL/DESY (Hamburg) and SLS (Villigen). †

Microbial Development Group, ‡Membrane Protein Crystallography, §Protein Modelling Group

of Instituto de Tecnologia Química e Biológica António Xavier, Universidade Nova de Lisboa, ITQB-UNL, 2780-157 Oeiras, Portugal ¤

Biotechnology Branch, Center for Disease Control and Prevention, Atlanta, GA 30333, USA 1

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

±

Page 2 of 43

These authors contributed equally to this work



Present addresses: CGF: Department of Microbiology and Immunology, Emory University

School of Medicine, Atlanta, Georgia 30322, United States of America; AI: Fundação para a Ciência e Tecnologia, Av. D. Carlos I, 126, 1249-074 Lisboa, Portugal ≈

Current address: Fundação para a Ciência e Tecnologia, Av. D. Carlos I, 126, 1249-074 Lisboa,

Portugal *MA: Phone: (+351) 214469747; Email: [email protected]; *AOH: Phone: (+351) 214411277; Email: [email protected] KEYWORDS: TGase, bacterial spore coat, NlpC/P60 endopeptidases, papain, cystamine RUNNING TITLE: Characterization of an ancient bacterial TGase

2

ACS Paragon Plus Environment

Page 3 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

ABBREVIATIONS TGase, transglutaminase; PCP, peptidoglycan cysteine endopeptidase; GST, glutathione Stransferase; MD, molecular dynamics.

3

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 43

ABSTRACT: Transglutaminases are best known for their ability to catalyze protein cross-linking reactions which impart chemical and physical resilience to cellular structures. Here, we report the crystal structure and characterization of Tgl, a transglutaminase from the bacterium Bacillus subtilis. Tgl is produced during sporulation and cross-links the surface of the highly resilient spore. Tgl-like proteins are only found in spore-forming bacteria of the Bacilli and Clostridia classes, indicating an ancient origin. Tgl is a single domain protein, produced in active form, and the smallest transglutaminase characterized to date. We show that Tgl shares structural similarity to bacterial cell wall endopeptidases and has an NlpC/P60 catalytic core, thought to represent the ancestral unit of the cysteine protease fold. We show that Tgl functions through a unique partially redundant catalytic dyad formed by Cys116, and Glu187 or Glu115. Strikingly, the catalytic Cys is insulated within a hydrophobic tunnel that transverses the molecule from side to side. The lack of similarity of Tgl to other transglutaminases together with its small size, suggests that an NlpC/P60 catalytic core and insulation of the active site during catalysis may be essential requirements for protein cross-linking.

4

ACS Paragon Plus Environment

Page 5 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Transglutaminases (TGases) catalyze transamidation reactions and are best known for their ability to crosslink proteins. Protein cross-linking results from the formation of stable and protease resistant ε-(γ-glutamyl)lysyl isopeptide bonds between protein-bound glutaminyl and lysyl residues, the Q and K substrates, respectively.1,2 Protein crosslinking is central to a variety of important processes including clotting phenomena, cell and tissue morphogenesis and development.1,2 Deregulation of TGase activity in humans contributes to a variety of severe pathologies including progressive fibrosis, neurodegenerative, autoimmune and infectious diseases, and illnesses related to malformations of the skin epidermis.3,4 The animal TGases, of which the human enzymes have been the most extensively studied, share a similar active site arrangement and acylation-deacylation pathway with the papain-like cysteine proteases5 (Figure 1). In general, TGases employ a charge-relay Cys-His-Asp catalytic triad with the reaction involving formation of a covalent acyl-enzyme intermediate (Fig 1b, step 3); during the reaction, a thiolate/imidazolium ion pair is formed between the catalytic Cys and His residues, the latter acting both as a proton donor and acceptor during catalysis (steps 2 and 4, in Figure 1B). An hydrogen bond between the Asp and His residues facilitates the correct orientation of the basic amino acid5-7 (Figure 1B). Adding to the mechanistic similarity, the catalytic core of TGases appears to derive from the minimal ancestral structural unit of the thiolprotease fold, whose closest current representative is found in the NlpC/P60 domain.8,9 This domain takes its name from the bacterial NlpC/P60 cell wall endopeptidases, which show a catalytic triad composed of a Cys, His, and a third polar residue.9,10 Because of the mechanistic and structural similarities it has been suggested that TGases evolved from an ancestral, papainlike protease.8,11

5

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 43

Reflecting their multiple functions and interactions in vivo and the need for tight control, the animal-like TGases are structurally complex and have intricate activation mechanisms.3,4,12 The animal-like TGases are activated by calcium and either produced in zymogenic form, bound by inhibitory subunits, and/or negatively modulated by GTP/GDP or ATP.2,12,13 Human TGase 2, for instance, a 80 kDa enzyme, adopts a compact, transamidase-inactive conformation in the absence of Ca2+.14 Activation of TGase 2 involves a large conformational rearrangement that exposes a hydrophobic tunnel leading to the active site.15 Concealment of the tunnel in the inactive form of the enzyme, as well as the presence of hydrophobic residues at its entry, is thought to reduce the potential for deamidation (hydrolytic) reactions.15 A tunnel is also seen in the calcium-activated forms of human TGase 3 and Factor XIIIa.16,17 In all cases, the active site is inaccessible from the outside in the inactive forms of the enzymes.14,16,18 Although an NlpC/P60 catalytic core together with the insulation of the active site from water may be essential characteristics of TGases, the structural complexity of the animal enzymes makes this inference difficult.19 Proteins related to the classical animal TGases, carrying a NlpC/P60 core, were detected in the genomes of several other groups of organisms, including bacteria.11 However, the NlpC/P60 domain defines a superfamily of proteins with distinct although related enzymatic activities.8 Thus, while one of the putative TGases, TgpA, a large multi-domain protein from Pseudomonas aeurginosa, indeed proved to be a TGase,20 several phage-associated proteins were shown to be proteases,21,22 and a protein from Bordetella bronchiseptica turned out to be involved in lipopolysaccharide maturation.23 Conversely, not all proteins with TGase activity are identifiable through sequence similarity as related to the animal-like TGases.24,25 GP42, for example, a large (80 kDa), multidomain, Ca2+-dependent TGase from a plant pathogen oomycete shows no

6

ACS Paragon Plus Environment

Page 7 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

sequence similarity to the classical TGases.26 Remarkably, GP42 also has a NlpC/P60 core and a Cys-His-Asp triad and seems related to a group of cysteine proteases from marine bacteria.26 The possible convergent evolution of GP42 raises the possibility that an NlpC/P60 core is an essential structural requirement for TGase activity.26 Two TGases of bacterial origin, which were also not detected by sequence similarity as being related to the animal-type enzymes,11 are much smaller in size. One, the MTG protein from Streptomyces mobaraensis, is the only bacterial TGase structurally characterized so far and appears to function through a Cys-Asp dyad.24,25 However, it was not reported as having a NlpC/P60 core24,25. While MTG is a 38 kDa, single domain, Ca2+-independent enzyme, it is still produced as a zymogen and shows a complex activation mechanism.27-29 The second is Tgl, found in a deeply rooted group of spore-forming microorganisms, which includes some Bacillus and Clostridium species, suggesting an ancient origin.30,31 The 28 kDa Tgl protein from B. subtilis has a well established role in cross-linking a multiprotein surface structure known as the coat that encases and protects the spore.32 Importantly, Tgl is produced in its active form and no factor (Ca2+, GTP or other) is known to control its activity.33 Here, we present the crystal structure of Tgl, the smallest TGase characterized to date, and show that the enzyme has the NlpC/P60 fold at its catalytic core and a unique, partially redundant, catalytic dyad. Moreover, the catalytic Cys116 is insulated within a hydrophobic tunnel which transverses the molecule form side to side. Tgl is structurally related to a group of bacterial cell wall endopeptidases. Together with the finding that MTG also possesses an NlpC/P60 core, the structure and functional characterization of Tgl herein reported supports the idea that, irrespective of their evolutionary history, all TGases have evolved from cysteine

7

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 43

proteases. Furthermore, the structure of Tgl offers insight into the minimal structural requirements for protein cross-linking by TGases.

EXPERIMENTAL PROCEDURES Protein production and purification. Point mutations were introduced in the tgl gene by PCR, creating pET30a(+) (Novagen) derivatives for the expression of Tgl, wild type or mutant forms. GST-SpoVID202 was expressed from pTC4934. All plasmids were introduced in E. coli BL21 (DE3) and the proteins were overproduced by a modified auto-induction method and the proteins purified by affinity chromatography. For activity assays, the purified proteins were dialyzed overnight against 0.1 M Tris-HCl, pH 8.0, using a 10 kDa cut-off membrane (SnakeSkin, from Pierce). For details on plasmid constructions, protein expression and purification refer to the Supporting Information. All primer sequences, plasmids and strains used in this study are indicated in Tables S1 and S2. Crystallization, X-ray data collection, phasing and refinement. Tgl was crystallized as previously described.33 Noteworthy, no Tgl crystals were formed at pH higher than 4.5-5. Transfer of crystals to solutions of higher pH led to their degradation, even with crystals crosslinked with glutaraldehyde. The Tgl structure was solved by single isomorphous replacement with the anomalous scattering (SIRAS), using the Tgl:Gd-derivative (Table S3). An initial model, comprising 501 residues in 8 chains, which belong to two molecules in the asymmetric unit, was built and improved. The refined structure of the Tgl:cysteamine complex was obtained from crystals soaked in a cystamine [2,2'-Dithiobis(ethylamine)] solution.

8

ACS Paragon Plus Environment

Page 9 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

The stereochemistry of the refined model was assessed with Procheck/MolProbity35,36 (Table S3). Structure representations were rendered with Pymol.37 For details see Supporting Information. Dansylcadaverine labeling assays. Amine incorporation activity was assessed by monitoring labeling of BSA (New England Biolabs) by dansylcadaverine (dansyl-cd, Fluka).38 Reaction mixtures containing Tgl (16 µM), BSA (60 µM) and dansyl-cd (0.5 mM) in Tris-HCl 0.1 M (pH 8.0) were incubated at 50º C, samples withdrawn along time and resolved on a 10% SDS-PAGE gel. Labeling of BSA by dansyl-cd was recorded on a UV transilluminator (Chemidoc XRS, Biorad) and quantified using ImageJ 1.37v.39 A sample collected following 120 min of reaction with wild type Tgl was loaded onto each gel for the normalization of the fluorescence signal among different assays. All Tgl forms were purified and assayed at least three times independently. Cross-linking assays. The wild type or mutant forms of Tgl (16 µM) were incubated with GST-SpoVID202 (60 µM) at 50º C in 0.1 M Tris-HCl, pH 8.0, and samples collected at different times for analysis on 10% SDS-PAGE gels. As controls, GST-SpoVID202 and Tgl were incubated alone, and samples taken at the beginning and at the end of the assays. All forms of Tgl were assayed independently a minimum of three times. Mass spectrometry. The high molecular weight species detected upon incubation of GSTSpoVID202 with Tglwt, TglE187A and TglH200A were analyzed by mass spectrometry and by Nterminal sequence (for the case of assays with wild type Tgl) where GST-SpoVID202 and Tgl (wild type or mutant forms) where positively identified (for details see Supporting Information). Molecular dynamics simulations. Molecular dynamics (MD) simulations were performed with the GROMACS package40, version 4.0.441, using the 53A6 force field42. Details of the

9

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 43

methodology used for the calculations of the protonation states and for the MD simulations are given in the Supporting Information. Accession Numbers. The atomic coordinates and the structure factor amplitudes of the Tgl crystal structure were deposited in the PDB under the accession numbers 4P8I (apo form) and 4PA5 (Tgl:cysteamine complex).

RESULTS Structure of Tgl. The X-ray structures of Tgl were obtained from crystals soaked with cystamine, a TGase inhibitor.43 Two X-ray structures of Tgl, in apo-form and bound to cysteamine (the reduced form of cystamine) were refined to 1.85 Å and 1.86 Å with R/Rfree values of 15.8/20.8% and 16.6/21.2%, respectively (relevant statistics are listed in Table S3). The apo and Tgl:cysteamine structures form dimers, which are very similar (RMSD of 0.29 Å for all Cα atoms). For simplicity, only the Tgl:cysteamine structure will be described in detail. The Tgl:cysteamine complex comprises 490 amino acid residues, a cysteamine in chain A (absent in chain B; Figure 2A-C), 4 PEG fragments, 2 sulfate ions, 1 citrate, 1 glycerol, 1 zinc ion and 204 water molecules. The Ramachandran plot shows 95.2% of the residues in the favored region and only 0.8% in the disallowed region.44 The electron density maps are generally of good quality, except for three disordered regions: residues 1-20 in both chains (Asp13-Glu15 not included in the model), and residues 180-184 and 219-224 in chain B. Each molecule has approximate dimensions of 60 x 50 x 30 Å and comprises three β-sheets (two-, four- and six-βstranded) and ten helices, of which three are 310 helices (Figure 2D). Cys116, essential for the activity of Tgl both in vivo and in vitro,45,46 is located at the Nterminus of a long helix, α6 (Figure 2D). In monomer A, the side-chain of Cys116 (Cys116A) is 10

ACS Paragon Plus Environment

Page 11 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

split into two conformations and additional electron density consistently appeared around the Sγ atom. This extra density was modeled as cysteamine (-S-CH2-CH2-NH2) covalently bound to conformer B and was refined to an occupancy of 88% (Figure 2A). For human TGase 2, under reducing conditions, cystamine (NH2-CH2-CH2-S-S-CH2-CH2-NH2) is readily reduced to two cysteamine (NH2-CH2-CH2-SH) molecules which then act as a K-donor substrate, competitively inhibiting the enzyme.43 In contrast, the structure deriving from inhibition of Tgl by cystamine is suggestive of a disulphide-exchange reaction with the catalytic Cys116, and evidence for an alternative mechanism of inhibition that may occur with many different TGases. In monomer B, Cys116B was modeled in a triple conformation (Figure 2B) with two nearby water molecules (not shown). These water molecules are within H-bonding distances to Glu187B and to one of two alternate conformations of the Glu115B side-chain. In monomer A, the second conformation of Glu115 is not present and a molecule of ethylene glycol occupies the corresponding space (Figure 2A). In all, the results suggest high flexibility around Cys116 and Glu115. The hydrophobic tunnel of Tgl and substrate accessibility. Strikingly, the Tgl structure shows a tunnel near the surface that crosses the molecule from side to side (Figure 3A) and which is not apparent in monomer A because of the covalent binding of cysteamine to Cys116 (not shown). The tunnel is about 15 Å long with its narrowest point (4.9 Å) approximately at the center (Figure 3B). The front and back entrances are 6.8 Å and 7.7 Å across, respectively (Figure 3C,D). The front side of the tunnel includes residues Phe72, Trp149, Tyr171 and Leu214 (Figure 3C) and shows a network of solvent molecules, including a PEG fragment, while the backside has fewer water molecules (not shown). A conspicuous feature of the backside is the close stacking of Trp184 with Phe217 (nearest distance is ~3.5 Å) (Figure 3D). Together with Phe69, 11

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 43

these aromatic residues are in parallel packing and create a hydrophobic environment to which Leu202 also contributes (Figure 3D). With the exception of Phe217, all of the aromatic residues at both tunnel entrances are highly conserved among Tgl orthologues as shown by a sequence Logo47 (Figure 3E and S1), and may contribute to the insulation of the active site from water. The left sidewall of the channel, as viewed from its front side entrance, includes Phe69, Phe72, Glu115, the catalytic Cys116, Ala117, Asp148 and Trp149 (Figure 3C,D). The right side comprises Tyr171, Gln183 to Asn188, His200 and Gly201 (Figure 3C,D), most of which are highly conserved (Figures 3E and S1). Some right wall residues show high thermal displacement parameters (B-factors), in particular Arg185, suggesting flexibility (Figures 2C and 3A). Flexibility at this region of the protein may be important for displacement of the ceiling of the tunnel, so that following catalysis the cross-linked protein product may be released and the enzyme regenerated.15,19 The catalytic core of Tgl adopts the NlpC/P60 papain-like fold. Proteins related to Tgl at the primary structure level can only be found in spore-forming bacteria of the Bacillus and Clostridium genus, and closely related organisms.30,31 However, a DALI search48 reveals structural similarity between Tgl and several proteins having in common the presence of an NlpC/P60 core. The highest hits were with the soluble domain (residues 37-162) of the Spr protein from Escherichia coli,10 and two homologous cyanobacterial proteins, NpPCP, from Nostoc punctiforme and AvPCP, from Anabaena variabilis9 (Figure 4A,B). All three proteins are peptidoglycan cysteine endopeptidases (PCPs) that exhibit Z scores around 7 and RMSD between 2.3-2.7 Å for the 188 (for NpPCP and AvPCP) or 126 (for Spr) residues aligned with Tgl. Structural elements shared by Tgl and the PCPs include helices α2 and α6, which are longer in Tgl, and 4 β-strands (β8 to β11) of an anti-parallel β-sheet element which is packed against α6

12

ACS Paragon Plus Environment

Page 13 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(Figure 4A,B). A long α-helix with the catalytic Cys residue at its N-terminus (Cys116 in Tgl), against which is packed a three-stranded anti-parallel β-sheet that contributes with additional catalytic residues, typifies the core NlpC/P60 fold.8 The core of the NlpC/P60 domain is thought to represent the ancestral, minimal fold of the papain-like cysteine proteases and TGases.8 In the papain-like and NlpC/P60 proteases, the core β-sheet element is part of a 5-stranded β-barrel, while in animal TGases this β-barrel is composed by 8 strands.8-10 The catalytic core of Tgl thus adopts the α+β fold of the core NlpC/P60 domain, with the β-barrel element consisting of 6 strands (β7-β-12; Figure 2D and 4A,B). Active site architecture. NpPCP and AvPCP, and presumably also Spr, use a Cys-His-His catalytic triad to cleave the peptidoglycan stem peptide.9,10 A closer inspection of the superimposed structures of Tgl, Spr, and NpPCP in the vicinity of the catalytic Cys shows that the position occupied by one of the catalytic His residues in the cell wall endopeptidases corresponds to Glu187 in Tgl, while the second His is close to His200 (Figure 4C). Both His200 (on a loop between strands β10 and β11), and Glu187 (on β9) are located within the six-stranded β-sheet of the NlpC/P60 core of Tgl (Figure 2D and 4A,B). Thus, the position of His200 and Glu187 in the structure seems to suggest that both are catalytic residues (but see below). We also superimposed the region around Cys116 of Tgl with the active site of MTG, the only other current structurally characterized microbial TGase,24,25 which as we show, also possesses a NlpC/P60 core (Figure S2), and human TGase 3, an archetypal animal TGase. In TGases, catalysis involves a Cys-His-Asp triad (for the animal enzymes) or a Cys-Asp dyad (for MTG) (Figure 1B). The relative position of the catalytic residues in MTG and TGase 3 results from a permutation, such that the His and Asp residues in MTG spatially correspond to the Asp and His residues of TGase 3, respectively (Figure 4D).24 This is also a functional permutation as Asp255

13

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 43

of MTG plays the role of the catalytic His in the animal-like TGases while His274 of MTG is not essential for catalysis.24,49 The superimposition of the catalytic Cys in Tgl (Cys116), MTG (Cys64) and TGase 3 (Cys272), places His200 of Tgl in the same relative position of His274 in MTG, raising the possibility that His200 is also not essential for catalysis. Furthermore, it also suggests that Glu187 in Tgl could be a catalytic residue: not only does it occupy the position equivalent to His330 in TGase 3 or Asp255 in MTG, but the nearest side-chain atoms of Glu187 are close (~4.3 Å) to the Sγ atom of the catalytic Cys116 residue (Figure 4D). Probing the active site of Tgl. To define the catalytic residues of Tgl, forms of the enzyme bearing single Ala substitutions of residues Cys116 (TglC116A), Glu187 (TglE187A) and His200 (TglH200A) were purified and tested in parallel with the wild type enzyme (Tglwt), for activity. In addition to protein cross-linking, a second transamidation reaction catalyzed by TGases involves a protein substrate and a primary amine present in simple molecules or in naturally occurring polyamines, which results in amine incorporation.1,2 In previous work, we have shown that BSA serves as a substrate for Tgl cross-linking and we reasoned that it could also be used in amine incorporation reactions.33 We first monitored the incorporation of the fluorescent K donor substrate dansylcadaverine (dansyl-cd) into bovine serum albumin (BSA), the Q substrate. Tglwt efficiently catalyzed the transfer of dansyl-cd to BSA, resulting in fluorescent BSA (Figure 5A), with labeling increasing linearly with time, during the first 45 min of the assay (Figure 5B). Importantly, cross-linking of BSA, auto-cross-linking of Tgl, or labeling of Tgl by dansyl-cd was not significant during the time of the assay (not shown). In confirmation and extension of earlier results45,46 the C116A substitution resulted in loss of enzymatic activity (Figure 5A). In contrast, labeling of BSA could be detected for TglH200A (albeit at low levels) and TglE187A, the latter showing about 60% of the wild type enzyme activity (Figure 5A,B).

14

ACS Paragon Plus Environment

Page 15 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Secondly, we assessed the ability of the various forms of Tgl to cross-link a glutathione Stransferase (GST) fusion protein. GST (about 30 kDa) can serve as a substrate for TGases.38,50 However, in screening a collection of proteins available in the laboratory, we found that a fusion of GST to the first 202 amino acids of coat protein SpoVID,34 was cross-linked by Tgl. The use of GST-SpoVID202, (50 kDa) increased the mass difference to Tgl (28 kDa), and allowed the enzyme and substrate to be unambiguously identified and their levels monitored on a Coomassiestained gel. Incubation of Tglwt, TglE187A or TglH200A with GST-SpoVID202 resulted in the formation of high molecular weight species over time (indicated by the parenthesis in Figure 5C), although at reduced levels for TglH200A. In contrast, these species were not detected upon incubation of GST-SpoVID202 with TglC116A or when Tgl or GST-SpoVID202 were incubated alone (Figure 5C,D). The more defined 75, 100 and 250 kDa bands detected after 180 min of incubation with Tglwt, TglE187A and TglH200A (Figure 5C, asterisks) were shown by mass spectrometry to contain sequences of both Tgl and GST-SpoVID202. While Tgl alone shows little auto-cross-linking activity under the conditions of the assay (Figure 5C, lanes with no GSTSpoVID202), these results suggest that Tglwt, TglE187A or TglH200A catalyze their cross-linking to GST-SpoVID202. In any event, formation of the 75, 100 and 250 kDa species is the result of Tgl activity. Thus, both the dansyl-cd incorporation and cross-linking assays show that unlike Cys116, Glu187 and His200 are not essential for catalysis. Interactions among residues near the catalytic Cys. Our activity assays suggest that the most likely partner and proton acceptor for Cys116 is His200. However this seems in contradiction with the apparent orientation of His200 in the crystal structure. As can be seen in Figure 4D, the His200 imidazole ring is not only too far from Sγ of Cys116 to allow proton abstraction (closest distance is ~ 5.7 Å), as its side chain is also oriented away from the active

15

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 43

site Cys (Nδ1 is H-bonded to a water molecule). Because the X-ray structures of Tgl were obtained at pH 4.0, outside the range for optimal enzyme activity,51 we reasoned that our crystallization conditions may have caused artifacts in the relative orientation of critical side chains in the active site region, specifically those of Glu187 and His200. Because we could not obtain crystals at a higher pH (see Experimental Procedures), molecular dynamics (MD) simulations were conducted at pH 7 (the MD simulations are described in detail in the Supporting Information). Our results indicate that the orientation of His200 detected in the crystal structure is a consequence of the low pH crystallization conditions and that an increase in pH can change the orientation of His200 towards the active site pocket. To gain further insight into the contribution of His200 and Glu187 to the activity of Tgl, we analyzed the hydrogen bonding interactions among the Cys116, Glu187 and His200 residues, as well as the correlations between all pairs of hydrogen bonds at pH 7.0 (Figure 6A; see also Tables S4 and S5). The results suggest that Cys116 and His200 do not interact through hydrogen bonding (Figure 6A), and therefore, His200 probably does not act as the proton acceptor for Cys116 (step 1 in Figure 1B). However, Glu187 forms a hydrogen bond with Cys116 with a high frequency (of about 78%) (Figure 6A). Furthermore, while Glu187 may interact with His200, the simultaneous interaction of Glu187 with both His200 and Cys116 appears unfavorable, as the Cys116-Glu187 and His200-Glu187 pairs show a negative correlation (Figure 6A). Therefore, and in spite of its presumed orientation towards the active site at pH 7.0, as suggested by our MD analysis (above), His200 does not appear to be a catalytic residue. This is consistent with the observation that TglH200A still retains activity in both amine incorporation and cross-linking assays (Figure 5A-C). In contrast, Glu187 appears to be a catalytic residue serving the role of the proton acceptor for Cys116. However, TglE187A still retains considerable activity (Figure 5A-C),

16

ACS Paragon Plus Environment

Page 17 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

raising the possibility that another residue, near the catalytic Cys116, can compensate for the absence of Glu187. A new catalytic residue. Our MD analysis of the interplay between Cys116 and its neighboring residues suggests that a second Glu residue, Glu115, may form a transient hydrogen bond with Cys116, with a frequency of about 20% (Figure 6A). As such, at pH 7, Glu115 is most likely deprotonated and may function as the proton acceptor for Cys116 (step 1 in Figure 1B). Furthermore, we found a strong negative correlation between the Cys116-Glu187 and the Cys116-Glu115 pairs, suggesting that the two acids compete for the catalytic Cys (Figure 6A). This could explain why TglE187A still shows enzymatic activity (Figure 5A-C). His200 and Glu115 also appear to form a hydrogen bond, but the simulations suggest that this is a rare event (frequency of 3%), especially when compared to formation of the His200-Glu187 salt bridge (Figure 6A; see also above). In addition, the simulations also suggest a negative correlation between the Cys116-Glu115 and His200-Glu115 pairs (Figure 6A). We infer that, as was also seen for Glu187, Glu115 does not tend to simultaneously interact with both Cys116 and His200. Finally, we also found strong positive correlations between the His200-Glu187 and Cys116Glu115 interactions (Figure 6A). Thus, the interaction between His200 and Glu187 reduces the Glu187-Cys116 interaction, so that Cys116 turns to the free Glu115. Similarly, the Cys116Glu187 and His200-Glu115 pairs are positively correlated (Table S5). It follows that the His200Glu115 interaction directs the catalytic Cys116 to the alternative Glu187. While strengthening the view that His200 is not a catalytic residue, these observations suggest that Tgl uses a catalytic dyad formed by either Cys116-Glu187 or Cys116-Glu115. A partially redundant catalytic dyad: non-reciprocal substitution of Glu187 by Glu115. The results presented in the preceding section suggested that Tgl has a catalytic dyad. This dyad

17

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 43

is however atypical, as the two acidic residues, Glu187 and Glu115, appear redundant. If so, then an E115A substitution would not be expected to eliminate enzyme activity. To test this prediction, purified TglE115A was tested for activity. Contrary to our expectations, however, the single E115A substitution rendered Tgl unable to detectably label BSA (Figure 5A) or to crosslink GST-SpoVID202 (Figure 5D). Thus, while Glu115 seems to replace Glu187, the reverse is not observed. In an attempt to understand this effect, we conducted MD simulations using in silico mutants (Figure 6B,C and S3D,E). The results show that in TglE187A the hydrogen bond between Cys116 and Glu115 becomes very likely, occurring with a frequency of about 80%, i.e., 4 times higher than in the wild type enzyme (Figure 6B). This is in line with the idea that in the absence of the Glu187 side chain, Cys116 readily interacts with Glu115. Also, in TglE187A, the frequency of the interaction between His200 and Glu115 is low (about 2%; Figure 6B and S3D) and, in fact, the average structures of the last 5 ns of the simulations with TglE187A show that His200 does not tend to interact with any other active site residue (Figure 6B and Table S4). Likewise, the simulations with the TglE115A mutant (Figure 6C and S3E) indicate that Cys116 tends to interact with the remaining acidic residue, Glu187, with the frequency of hydrogen bonding between Cys116 and Glu187 increasing relative to the wild type (from 78.3 to 93.2%). Importantly however, and unlike the results for the wild type enzyme or for TglE187A, the absence of Glu115 additionally leads to the frequent interaction between His200 and Glu187 (frequency of 93%; Figure 6C and S3E). Thus, under these conditions, the Glu187-His200 interaction is stable. The once negative correlation between the Cys116-Glu187 and His200-Glu187 pairs is now positive, and both Cys116 and His200 appear to interact simultaneously with Glu187, which acts as a proton acceptor for both (Figure 6C). Considering that in TglE115A, Glu187 is the proton acceptor for Cys116 then the simultaneous interaction with His200 would decrease its ability for

18

ACS Paragon Plus Environment

Page 19 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

proton abstraction and the production of a thiolate ion (step 1 in Figure 1B). The result would be the decreased reactivity of Cys116. This possibly explains the strongly reduced activity of TglE115A and the partial redundancy of the Tgl Cys116-Glu187 or Cys116-Glu115 catalytic dyad.

DISCUSSION Here, we present the crystal structure and functional characterization of Tgl, a 28 kDa sporulation-specific bacterial TGase. Sporulation emerged at the basis of the Firmicutes Phylum, some 3 billion years ago.30 A much larger TGase (YebA, with a predicted molecular weight of 75 kDa), related to the animal-type enzymes, is also detected in B. subtilis.11 Thus, it seems possible that Tgl emerged as a sporulation-specific TGase dedicated to the cross-linking of the spore surface.46,52-54 Although the evolutionary history of Tgl is presently unknown, its structural similarity to a group of bacterial cell wall endopepetidases suggests that regardless of their origin, all TGases evolved from cysteine proteases. Moreover, it shows that the adoption of the NlpC/P60 core by TGases is ancient. Another finding of this study is that the catalytic center of Tgl is located within a 15 Å-long tunnel that crosses the molecule from side to side, with both entrances surrounded by hydrophobic residues (Figure 3B-D). This tunnel is a second feature that Tgl shares with the active form of animal TGases. The Tgl tunnel appears wide enough (c.a. 6 Å) to accommodate a peptide chain backbone bearing a reactive Q or K residue. Presumably, this tunnel contributes to waterproofing the active site favoring transamidation over deamidation reactions. The insulation of the catalytic Cys within a hydrophobic tunnel that channels the Q and K substrates to the active site has been proposed as an essential structural requirement for protein cross-linking by

19

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 43

TGases.19 This requirement is fulfilled in the active forms of human TGase 2 and 3 and Factor XIIIa15-17 and may be a conserved feature of the animal-type enzymes.4 The finding of a tunnel in Tgl, suggests that the adoption of this structural feature was also an early event. Strikingly, in MTG and GP42, the active site is close to the surface of the molecule.24,26 However, the prosequence that covers the active site cleft of MTG, positions two Tyr residues on top of the catalytic residues25 and in GP42, several aromatic residues are located above the catalytic Cys.26 It is tempting to suggest that binding of MTG or GP42 substrates may exclude water from their active sites during catalysis. In any event, the animal-like enzymes,8 as well as GP42,26 MTG (see Figure S2), and the smaller sized Tgl, all show an NlpC/P60 core. This suggests that an NlpC/P60 core, together with a mechanism for active site insulation, may be universal requirements for protein cross-linking. Both entries of the tunnel are most likely docking sites for the Q and K substrates of Tgl. Interestingly while many of the residues at both entrances of the tunnel are conserved among Tgl orthologues, some stretches are highly variable (Figure 3E). It is possible that this variability represents specificity towards different substrates. We do not presently know which sides of the tunnel are engaged by the Q and K substrates. However, waterproofing of the K acceptor side is important to reduce the potential for hydrolytic reactions, as shown for human TGase 2.15 Thus, the K acceptor side of Tgl may correspond to the more hydrophobic “back” entrance of the tunnel (Figure 3E). His200, at the “back” entrance of the Tgl tunnel does not seem to be a catalytic residue (see above), as the simultaneous interactions among Cys116, Glu187 and His200 are unfavorable. Rather, His200 may be involved in substrate recognition, contributing in that way to the overall activity of the enzyme. However, the initial deprotonation of the acylacceptor (K) substrate, which in solution is likely to exist predominantly in its protonated form,

20

ACS Paragon Plus Environment

Page 21 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

has been proposed to be an important step in the reaction mechanism of TGases.17,55 In factor XIII and in human TG2, His residues outside the catalytic triad appear to facilitate the nucleophilic attack by the K substrate on the acyl-enzyme intermediate.17,55 It is tempting to suggest that His200 plays a similar role, perhaps coupling binding of the K substrate to its deprotonation. An Ala substitution of His274 in MTG, which occupies the same relative position of His200 in Tgl (Figure 4D), also reduces but does not eliminate the activity of the enzyme.24,49 In MTG, His274 was proposed to be involved in substrate interactions.49 However, it could also be involved in the initial deprotonation of the K substrate, as suggested above for Tgl. A third significant finding is that Tgl has a distinctive catalytic center. While Glu187 appears to be the primary proton acceptor for Cys116, a second acidic residue, Glu115, in the close vicinity of Cys116, can substitute for Glu187, at least when the latter is substituted by an Ala. The reciprocal substitution (E115A) abrogates enzyme activity, possibly because in the mutant, as suggested by the MD simulations, Glu187 interacts with equal frequency with both Cys116 and His200, reducing its capacity to deprotonate Cys116. While Glu115 (but not Glu187) is strictly conserved among Tgl orthologues (Figure 3E), further studies are needed to elucidate its role in catalysis. Both the MD simulations and the activity assays converge to indicate that Tgl has a catalytic dyad. Why the two bacterial TGases, MTG and Tgl, employ a catalytic dyad,24 whereas the animal-type enzymes use a triad is presently unclear. However, it has been suggested that in animal-like TGases in addition to the classical Cys-His-Asp triad, a Cys-His catalytic dyad may also be involved in catalysis.17 Interestingly, papain also seems to function through a dual catalytic Cys26-His159-Asp158/Asn175 triad.56 The dual partial redundant catalytic dyad of Tgl (Cys116/Glu187 or Glu115) implies that catalysis may take place from one or two distinct steric positions. Perhaps in Tgl different physiological substrates are handled in

21

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 43

distinct ways. Variations in the spatial arrangement of the catalytic residues, such as the permutations seen in MTG and Tgl (Figure 3), may in fact reflect enzyme-substrate interactions. An extreme example may be found in GP42, in which the His of the Cys-His-Asp triad is located in the same helix, part of the NlpC/P60 core, that carries the catalytic Cys, while the β-barrel element contributes with aromatic residues possibly involved in insulation of the catalytic Cys (as mentioned above) and substrate binding. Rearrangement of the triad was proposed to have occurred during evolution so that aromatic residues required for substrate specificity could be accommodated.26 Substrate interactions may also induce important functional rearrangements of the active site region. Our MD simulations also show that unlike what has been proposed for the animal TGases, where the catalytic Cys appears as a thiolate ion in the initial step of the reaction mechanism (Figure 1B, step 1), in Tgl the Cys appears protonated even at high pH values (Supporting Information). It is thus possible that in Tgl binding of a substrate will induce modifications in the pKa of the catalytic Cys and the formation of the thiolate ion will be concomitant with the initial nucleophilic attack on the Q substrate and the transfer of the proton to Glu187 or Glu115. TGases catalyze a variety of reactions of which protein cross-linking is the most notorious one. They play important roles in cell and tissue morphogenesis, and participate in a variety of processes, from clotting phenomena to development. Proteins homologous to the animal-type TGases, of which the human enzymes are the most extensively studied, show a broad phylogenetic distribution. They are structurally and mechanistically related to the papain-like cysteine proteases and accordingly, show an NlpC/P60-type catalytic core, though to represent the minimal structural unit of the thiol protease fold. The TGases are large multidomain

22

ACS Paragon Plus Environment

Page 23 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

enzymes, and/or with complex activation mechanisms, making the identification of the essential requirements for protein cross-linking difficult. The bacterial TGase, Tgl, a single domain enzyme, is the smallest TGase characterized to date. It most likely emerged at the basis of the Firmicutes Phylum, some 3 billion years ago, as an enzyme dedicated to the cross-linking of the spore coat. Tgl is structurally related to a family of bacterial cell wall endopeptidases, has an NlpC/P60 core, and a hydrophobic tunnel that insulates the active site. These features may thus be essential requirements for protein cross-linking. Moreover, it has a partially redundant catalytic dyad, reminiscent of papain. The properties of Tgl lend support to the notion that all TGases have evolved from ancestral thiol proteases.

AUTHOR CONTRIBUTIONS DP collected the diffraction data; DP, JAB and MA analyzed the diffraction data and determined the structure; CGF and AI performed mutagenic studies; CGF conducted the functional characterization of Tgl and its mutants; DL and CMS performed the molecular dynamics studies; CGF, DP, MA, AOH and MAC conceived the experimental design and interpreted the data. CGF, DP, DL, JAB, CMS, MA, AOH and MAC prepared the manuscript.

AKNOWLEDGMENTS We thank Pedro Matias (ITQB) for his help with data collection. We also acknowledge support from the European Synchrotron Radiation Facility (ESRF, Grenoble).

23

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 43

Supporting Information available. Additional experimental procedures, detailed description of the molecular dynamics simulations methods and results, ClustalW alignmentof Tgl orthologues, evidence of the NlpC/P60 core in MTG, data collection, phasing and refinement statistics of the Tgl structure, interaction frequency and correlation analysis among active site residues. This material is available free of charge via the Internet at http://pubs.acs.org.

REFERENCES (1) Griffin, M., Casadio, R., and Bergamini, C. M. (2002) Transglutaminases: nature's biological glues, The Biochemical journal 368, 377-396. (2) Lorand, L., and Graham, R. M. (2003) Transglutaminases: crosslinking enzymes with pleiotropic functions, Nature reviews 4, 140-156. (3) Kim, S. Y., Jeitner, T. M., and Steinert, P. M. (2002) Transglutaminases in disease, Neurochemistry international 40, 85-103. (4) Iismaa, S. E., Mearns, B. M., Lorand, L., and Graham, R. M. (2009) Transglutaminases and disease: lessons from genetically engineered mouse models and inherited disorders, Physiological reviews 89, 991-1023. (5) Pedersen, L. C., Yee, V. C., Bishop, P. D., Le Trong, I., Teller, D. C., and Stenkamp, R. E. (1994) Transglutaminase factor XIII uses proteinase-like catalytic triad to crosslink macromolecules, Protein Sci 3, 1131-1135.

24

ACS Paragon Plus Environment

Page 25 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(6) Ahvazi, B., Boeshans, K. M., Idler, W., Baxa, U., and Steinert, P. M. (2003) Roles of calcium ions in the activation and activity of the transglutaminase 3 enzyme, The Journal of biological chemistry 278, 23834-23841. (7) Boeshans, K. M., Mueser, T. C., and Ahvazi, B. (2007) A three-dimensional model of the human transglutaminase 1: insights into the understanding of lamellar ichthyosis, Journal of molecular modeling 13, 233-246. (8) Anantharaman, V., and Aravind, L. (2003) Evolutionary history, structural features and biochemical diversity of the NlpC/P60 superfamily of enzymes, Genome biology 4, R11. (9) Xu, Q., Sudek, S., McMullan, D., Miller, M. D., Geierstanger, B., Jones, D. H., Krishna, S. S., Spraggon, G., Bursalay, B., Abdubek, P., Acosta, C., Ambing, E., Astakhova, T., Axelrod, H. L., Carlton, D., Caruthers, J., Chiu, H. J., Clayton, T., Deller, M. C., Duan, L., Elias, Y., Elsliger, M. A., Feuerhelm, J., Grzechnik, S. K., Hale, J., Won Han, G., Haugen, J., Jaroszewski, L., Jin, K. K., Klock, H. E., Knuth, M. W., Kozbial, P., Kumar, A., Marciano, D., Morse, A. T., Nigoghossian, E., Okach, L., Oommachen, S., Paulsen, J., Reyes, R., Rife, C. L., Trout, C. V., van den Bedem, H., Weekes, D., White, A., Wolf, G., Zubieta, C., Hodgson, K. O., Wooley, J., Deacon, A. M., Godzik, A., Lesley, S. A., and Wilson, I. A. (2009) Structural basis of murein peptide specificity of a gamma-D-glutamyl-l-diamino acid endopeptidase, Structure 17, 303-313. (10) Aramini, J. M., Rossi, P., Huang, Y. J., Zhao, L., Jiang, M., Maglaqui, M., Xiao, R., Locke, J., Nair, R., Rost, B., Acton, T. B., Inouye, M., and Montelione, G. T. (2008) Solution NMR structure of the NlpC/P60 domain of lipoprotein Spr from Escherichia coli: structural evidence for a novel cysteine peptidase catalytic triad, Biochemistry 47, 9715-9717.

25

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 43

(11) Makarova, K. S., Aravind, L., and Koonin, E. V. (1999) A superfamily of archaeal, bacterial, and eukaryotic proteins homologous to animal transglutaminases, Protein Sci 8, 17141719. (12) Gundemir, S., Colak, G., Tucholski, J., and Johnson, G. V. W. (2012) Transglutaminase 2: A molecular Swiss army knife, Biochimica et Biophysica Acta (BBA) - Molecular Cell Research 1823, 406-419. (13) Klöck, C., and Khosla, C. (2012) Regulation of the activities of the mammalian transglutaminase family of enzymes, Protein Science 21, 1781-1791. (14) Liu, S., Cerione, R. A., and Clardy, J. (2002) Structural basis for the guanine nucleotidebinding activity of tissue transglutaminase and its regulation of transamidation activity, Proceedings of the National Academy of Sciences of the United States of America 99, 2743-2747. (15) Pinkas, D. M., Strop, P., Brunger, A. T., and Khosla, C. (2007) Transglutaminase 2 undergoes a large conformational change upon activation, PLoS biology 5, e327. (16) Ahvazi, B., Kim, H. C., Kee, S. H., Nemes, Z., and Steinert, P. M. (2002) Threedimensional structure of the human transglutaminase 3 enzyme: binding of calcium ions changes structure for activation, The EMBO journal 21, 2055-2067. (17) Stieler, M., Weber, J., Hils, M., Kolb, P., Heine, A., Büchold, C., Pasternack, R., and Klebe, G. (2013) Structure of Active Coagulation Factor XIII Triggered by Calcium Binding: Basis for the Design of Next-Generation Anticoagulants, Angewandte Chemie International Edition 52, 11930-11934.

26

ACS Paragon Plus Environment

Page 27 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(18) Yee, V. C., Pedersen, L. C., Le Trong, I., Bishop, P. D., Stenkamp, R. E., and Teller, D. C. (1994) Three-dimensional structure of a transglutaminase: human blood coagulation factor XIII, Proceedings of the National Academy of Sciences of the United States of America 91, 72967300. (19) Nemes, Z., Petrovski, G., Csosz, E., and Fesus, L. (2005) Structure-function relationships of transglutaminases--a contemporary view, Progress in experimental tumor research. Fortschritte der experimentellen Tumorforschung 38, 19-36. (20) Milani, A., Vecchietti, D., Rusmini, R., and Bertoni, G. (2012) TgpA, a protein with a eukaryotic-like transglutaminase domain, plays a critical role in the viability of Pseudomonas aeruginosa, PLoS One 7, e50323. (21) Luo, Y., Pfister, P., Leisinger, T., and Wasserfallen, A. (2002) Pseudomurein endoisopeptidases PeiW and PeiP, two moderately related members of a novel family of proteases produced in Methanothermobacter strains, FEMS microbiology letters 208, 47-51. (22) Pfister, P., Wasserfallen, A., Stettler, R., and Leisinger, T. (1998) Molecular analysis of Methanobacterium phage psiM2, Molecular microbiology 30, 233-244. (23) King, J. D., Vinogradov, E., Preston, A., Li, J., and Maskell, D. J. (2009) Post-assembly modification of Bordetella bronchiseptica O polysaccharide by a novel periplasmic enzyme encoded by wbmE, The Journal of biological chemistry 284, 1474-1483. (24) Kashiwagi, T., Yokoyama, K., Ishikawa, K., Ono, K., Ejima, D., Matsui, H., and Suzuki, E. (2002) Crystal structure of microbial transglutaminase from Streptoverticillium mobaraense, The Journal of biological chemistry 277, 44252-44260.

27

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 43

(25) Yang, M. T., Chang, C. H., Wang, J. M., Wu, T. K., Wang, Y. K., Chang, C. Y., and Li, T. T. (2010) Crystal Structure and Inhibition Studies of Transglutaminase from Streptomyces mobaraense, Journal of Biological Chemistry 286, 7301-7307. (26) Reiss, K., Kirchner, E., Gijzen, M., Zocher, G., Loffelhardt, B., Nurnberger, T., Stehle, T., and Brunner, F. (2011) Structural and Phylogenetic Analyses of the GP42 Transglutaminase from Phytophthora sojae Reveal an Evolutionary Relationship between Oomycetes and Marine Vibrio Bacteria, Journal of Biological Chemistry 286, 42585-42593. (27) Pasternack, R., Dorsch, S., Otterbach, J. T., Robenek, I. R., Wolf, S., and Fuchsbauer, H. L. (1998) Bacterial pro-transglutaminase from Streptoverticillium mobaraense-purification, characterisation and sequence of the zymogen, European journal of biochemistry / FEBS 257, 570-576. (28) Zotzel, J., Keller, P., and Fuchsbauer, H. L. (2003) Transglutaminase from Streptomyces mobaraensis is activated by an endogenous metalloprotease, European journal of biochemistry / FEBS 270, 3214-3222. (29) Zotzel, J., Pasternack, R., Pelzer, C., Ziegert, D., Mainusch, M., and Fuchsbauer, H. L. (2003) Activated transglutaminase from Streptomyces mobaraensis is processed by a tripeptidyl aminopeptidase in the final step, European journal of biochemistry / FEBS 270, 4149-4155. (30) Abecasis, A. B., Serrano, M., Alves, R., Quintais, L., Pereira-Leal, J. B., and Henriques, A. O. (2013) A genomic signature and the identification of new sporulation genes, Journal of bacteriology 195, 2101-2115.

28

ACS Paragon Plus Environment

Page 29 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(31) Galperin, M. Y., Mekhedov, S. L., Puigbo, P., Smirnov, S., Wolf, Y. I., and Rigden, D. J. (2012) Genomic determinants of sporulation in Bacilli and Clostridia: towards the minimal set of sporulation-specific genes, Environmental microbiology 14, 2870-2890. (32) Henriques, A. O., and Moran, C. P., Jr. (2007) Structure, assembly, and function of the spore surface layers, Annual review of microbiology 61, 555-588. (33) Placido, D., Fernandes, C. G., Isidro, A., Carrondo, M. A., Henriques, A. O., and Archer, M. (2008) Auto-induction and purification of a Bacillus subtilis transglutaminase (Tgl) and its preliminary crystallographic characterization, Protein expression and purification 59, 1-8. (34) Costa, T., Isidro, A. L., Moran, C. P., Jr., and Henriques, A. O. (2006) Interaction between coat morphogenetic proteins SafA and SpoVID, Journal of bacteriology 188, 7731-7741. (35) Chen, V. B., Arendall, W. B., 3rd, Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2010) MolProbity: all-atom structure validation for macromolecular crystallography, Acta Crystallogr D Biol Crystallogr 66, 12-21. (36) Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) PROCHECK: a program to check the stereochemical quality of protein structures, J. Appl. Cryst. 26, 283-291. (37) DeLano. (2002) The PyMOL Molecular Graphics System Palo Alto, CA, USA. (38) Sugimura, Y., Hosono, M., Wada, F., Yoshimura, T., Maki, M., and Hitomi, K. (2006) Screening for the preferred substrate sequence of transglutaminase using a phage-displayed

29

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 43

peptide library: identification of peptide substrates for TGase 2 and Factor XIIIa, The Journal of biological chemistry 281, 17699-17706. (39) Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012) NIH Image to ImageJ: 25 years of image analysis, Nature methods 9, 671-675. (40) Berendsen, H. J. C., van der Spoel, D., and van Drunen, R. (1995) GROMACS: A message-passing

parallel

molecular

dynamics

implementation,

Computer

Physics

Communications 91, 43-56. (41) Hess, B., Kutzner, C., van der Spoel, D. & Lindahl, E. . (2008) GROMACS 4: Algorithms for highly efficient, load-balanced, and scalable molecular simulation, J. Chem. Theory Comput. 4, 435-447. (42) Oostenbrink, C., Villa, A., Mark, A. E., and van Gunsteren, W. F. (2004) A biomolecular force field based on the free enthalpy of hydration and solvation: the GROMOS force-field parameter sets 53A5 and 53A6, Journal of computational chemistry 25, 1656-1676. (43) Jeitner, T. M., Delikatny, E. J., Ahlqvist, J., Capper, H., and Cooper, A. J. (2005) Mechanism for the inhibition of transglutaminase 2 by cystamine, Biochemical pharmacology 69, 961-970. (44) Lovell, S. C., Davis, I. W., Arendall, W. B., 3rd, de Bakker, P. I., Word, J. M., Prisant, M. G., Richardson, J. S., and Richardson, D. C. (2003) Structure validation by Calpha geometry: phi,psi and Cbeta deviation, Proteins 50, 437-450.

30

ACS Paragon Plus Environment

Page 31 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(45) Kobayashi, K., Hashiguchi, K., Yokozeki, K., and Yamanaka, S. (1998) Molecular cloning of the transglutaminase gene from Bacillus subtilis and its expression in Escherichia coli, Bioscience, biotechnology, and biochemistry 62, 1109-1114. (46) Zilhao, R., Isticato, R., Martins, L. O., Steil, L., Volker, U., Ricca, E., Moran, C. P., Jr., and Henriques, A. O. (2005) Assembly and function of a spore coat-associated transglutaminase of Bacillus subtilis, Journal of bacteriology 187, 7753-7764. (47) Crooks, G. E., Hon, G., Chandonia, J. M., and Brenner, S. E. (2004) WebLogo: a sequence logo generator, Genome research 14, 1188-1190. (48) Holm, L., and Sander, C. (1993) Protein structure comparison by alignment of distance matrices, J Mol Biol 233, 123-138. (49) Tagami, U., Shimba, N., Nakamura, M., Yokoyama, K. i., Suzuki, E. i., and Hirokawa, T. (2009) Substrate specificity of microbial transglutaminase as revealed by three-dimensional docking simulation and mutagenesis, Protein Engineering Design and Selection 22, 747-752. (50) Piredda, L., Farrace, M. G., Lo Bello, M., Malorni, W., Melino, G., Petruzzelli, R., and Piacentini, M. (1999) Identification of 'tissue' transglutaminase binding proteins in neural cells committed to apoptosis, Faseb J 13, 355-364. (51) Kobayashi, K., Suzuki, S. I., Izawa, Y., Miwa, K., and Yamanaka, S. (1998) Transglutaminase in sporulating cells of Bacillus subtilis, J Gen Appl Microbiol 44, 85-91. (52) Monroe, A., and Setlow, P. (2006) Localization of the transglutaminase cross-linking sites in the Bacillus subtilis spore coat protein GerQ, Journal of bacteriology 188, 7609-7616.

31

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 43

(53) Ragkousi, K., and Setlow, P. (2004) Transglutaminase-mediated cross-linking of GerQ in the coats of Bacillus subtilis spores, Journal of bacteriology 186, 5567-5575. (54) Kuwana, R., Okuda, N., Takamatsu, H., and Watabe, K. (2006) Modification of GerQ reveals a functional relationship between Tgl and YabG in the coat of Bacillus subtilis spores, Journal of biochemistry 139, 887-901. (55) Keillor, J. W., Clouthier, C. M., Apperley, K. Y. P., Akbar, A., and Mulani, A. (2014) Acyl transfer mechanisms of tissue transglutaminase, Bioorganic Chemistry 57, 186-197. (56) Wang, J., Xiang, Y. F., and Lim, C. (1994) The double catalytic triad, Cys25-His159Asp158 and Cys25-His159-Asn175, in papain catalysis: role of Asp158 and Asn175, Protein engineering 7, 75-82.

FIGURE LEGENDS Figure 1. Proposed reaction mechanism for the animal TGases. (A) Superimposition of the catalytic residues of human TGase 3 (PDB entry 1L9M; orange) and papain (1PPN; blue). Note that only the Asn175 of the dual Cys26-His159-Asp158 or Cys26-His159-Asn175 catalytic triad of papain is represented56. (B) Reaction mechanism of TGases proposed on the basis of the similarity of their catalytic core to that of papain-like proteases. Gln and Lys refer to proteinbound Gln and Lys residues. The reaction mechanism involves 6 different steps: (1) the thiolate ion nucleophilically attacks the Gln side chain (the acyl donor or Q substrate), leading to the formation of an oxyanion intermediate (2); then, the regeneration of the carbonyl group of the tetrahedral intermediate leads to the release of ammonia and to the formation of the acyl-enzyme

32

ACS Paragon Plus Environment

Page 33 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

intermediate (3); the catalytic His removes a proton from the primary amine of the Lys residue (4) leading to the nucleophilic attack of the Lys residue on the carbonyl group of the acylenzyme intermediate formed in 3, and to the formation of a second tetrahedral intermediate (5); as the oxyanion intermediate regenerates the carbonyl group, the thioester bond is cleaved, regenerating the initial catalytic core and releasing the final, cross-linked, protein product (6). Figure 2. Active site and overall structure of Tgl. 2Fo–Fc electron density maps at 1.2 σ contour around relevant residues (Glu115, Cys116, Glu187, and His200) for monomers A (A) and B (B), of the Tgl:cysteamine complex, respectively (see also Table S3). Monomer A shows a cysteamine (Cyst) covalently bound to the Sγ atom of the conformer B of Cys116 and a molecule of ethylene glycol (EG) near the side-chain of Glu115. In monomer B, the side chain of Glu115 displays two conformations and Cys116 was modeled in three different conformations. Color code: O- red, N- blue and S- green, and C- yellow for monomer A and magenta for monomer B (the same color code is followed throughout the text for Tgl). (C) Superposition of monomers A and B. Note that Gln183 and Arg185 show slightly different conformations in both monomers. (D) Overall fold of Tgl; secondary structural elements are indicated as well as the position of Cys116 (black arrow). Figure 3. The Tgl tunnel. (A) Molecular surface representation of Tgl (same orientation as in Figure 2D), color-coded according to the thermal displacement parameters (B factors; from blue, lower, to red, higher). The region around the Tgl tunnel is encircled. (B) Mesh representation of the cavity in the active site region, viewed as a cross section from the top of the Tgl molecule (residues forming the ceiling of the tunnel are not represented for clarity). (C) Zoomed view of the front entry of the hydrophobic tunnel. (D) Zoomed view of the back entry of the tunnel (180º rotation around the y axis, relative to C). In B, C and D, the distances between relevant residues, 33

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 43

shown in stick representation, are indicated. (E) Sequence Logo47 of the blocks of residues that form the left and right wall of the tunnel (numbered according to the Tgl sequence), for 40 selected Tgl orthologues (see Figure S1). Residues, in singe letter notation, are coloured coded as follows: M, L, I, V, F, W, in grey; A, G, T, P, in orange; E, D, R, K, Q, N, in blue; C, green. The residues represented in panels c and d are highlighted in red. Figure 4. The catalytic core of Tgl has an NlpC/P60-like fold. (A) Structure-based amino acid sequence alignment of Tgl, NpPCP (PDB entry 2EVR), AvPCP (2HBW) and Spr (2KIG), generated with DALI48. Helices (red) and β-strands (blue) refer to the Tgl structure. Residues aligned in all four structures are shown in capital letters and highlighted in grey. The secondary structure elements corresponding to the NlpC/P60 core are marked with brown lines below the alignment. Actual or presumptive catalytic residues are shown in white against a dark red background. (B) Structural superposition of Tgl (in yellow) with NpPCP (cyan). NpPCP (as well as AvPCP) contains an N-terminal SH3-like domain (as indicated), which is absent in Tgl. The figure highlights the conserved core of the NlpC/P60 domain (strands β8-β11 in Tgl, forming a β-sheet packed against α2 and α6; only the clearly visible elements are labeled). The active site region is circled. Superimposition of the putative catalytic residues of Tgl (brown numbering) with: (C) NpPCP (cyan) and Spr (gray), and (D) TGase 3 (PDB entry 1L9M; orange) and MTG (1IU4; green). For the animal-type TGases (e.g., TGase 3), the arrangement of the catalytic residues (Cys-His-Asp) is very similar. The inversion of the His and Asp/Glu residues of MTG/Tgl when compared to animal TGases is illustrated. MTG shows a catalytic Cys-Asp dyad instead of the typical triad (see also Figure S2). Figure 5. Activity assays with wild type and mutant forms of Tgl. (A) Enzyme activity was monitored for Tglwt and for the TglC116A, TglE187A, TglH200A and TglE115A forms by the 34

ACS Paragon Plus Environment

Page 35 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

incorporation of fluorescent dansylcadaverine into BSA, as detected under UV light (BSA*) after SDS-PAGE of the reaction products. The top rows of each panel represent samples taken at the indicated reaction times (in min). The middle and bottom rows show the regions of the same gel, stained with Coomassie. The fluorescence signal was normalized using the level of BSA* detected after 120 min of incubation with Tglwt. (B) The normalized fluorescence was plotted for the various forms of the enzyme, except for TglC116A and TglE115A for which no labeled BSA could be detected; represented are the mean and standard deviation values for each time point. (C) and (D) Cross-linking assays of GST-SpoVID202 by Tglwt, TglH200A or TglE187A (C) and TglE115A or TglC116A (D). GST-SpoVID202 was incubated with the different forms of Tgl for the indicated times (in min), after which the reaction mixture was resolved by SDS-PAGE and the gels stained with Coomassie. The parenthesis indicates the region of the gel where cross-linked products are found (absent from control reactions where GST-SpoVID202, Tglwt or the mutant enzymes were incubated alone). The bands labeled with an asterisk were subjected to mass spectrometry and N-terminal sequence analysis. The slight decrease in the level of Tgl following 180 min of incubation, especially in the case of TglH200A or TglE115A, is likely due to precipitation as it proved to be independent of enzymatic activity, and was not seen in all the replicas of the assays. Fig. 6. Behavior of Tglwt, TglE187A and TglE115A in MD simulations. Two molecules were observed in the final model of the X-ray structure and both were simulated, named monomer A and B and 3 simulations were conducted for each monomer (see also Figure S3). Representation of the MD simulations for Tglwt (A), TglE187A (B) and TglE115A (C). In the left hand side of each panel is represented the average structure of the active site of Tgl during the last 5 ns of simulation. The right side of each panel shows a schematic representation of the interactions

35

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 43

(black dotted lines) formed among the residues studied and their frequency (numbers above dotted lines) and correspondent negative (red lines) or positive (green lines) correlations. For simplicity, the interaction between His200 and Glu187/Glu115 is depicted as occurring through the Nε2 proton of His200; however the interaction was also detected via the Nδ1 proton of His200, as seen on the average structure on the left side (see also Tables S4 and S5). From the average structure of the active site of Tgl it is clear that His200 does not interact with Cys116. Furthermore, in Tglwt (A) and TglE187A (B) the simultaneous interaction of Glu187 or Glu115 with both Cys116 and His200 is unfavorable unlike what is seen with TglE115A (C).

36

ACS Paragon Plus Environment

Page 37 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

FOR TABLE OF CONTENTS USE ONLY

37

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

181x113mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 38 of 43

Page 39 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

165x169mm (300 x 300 DPI)

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

219x166mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 40 of 43

Page 41 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

225x173mm (300 x 300 DPI)

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

224x173mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 42 of 43

Page 43 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

189x176mm (300 x 300 DPI)

ACS Paragon Plus Environment